The discharge of a river is one of the most important parameters of the hydraulic and hydrological studies such as hydropower production, canal design, irrigation, basin management. Each basin has different climate and geological characteristics which influence the regional infiltration capacity and runoff. The aim of this study is to estimate the average annual flow rates of ungauged locations on the Tigris River Basin. In total, eleven machine learning methods were applied to the long-term average annual discharge and the drainage area data of 34 flow measurement stations (FMS). Among all methods employed here, the conventional regression analysis was found to be the most successful test with a correlation coefficient (R2 value) of 0.96. The equation of the best fitted linear line represents the relationship between the drainage area and the discharge. The results of this study are expected to enable the prediction of the average annual flow rate of any sub-basin of the Tigris River.
Key words: Average annual discharge, the Tigris River, regression analysis, machine learning methods, ungauged basin.